Self-Aligned Concave Curve: Illumination Enhancement for Unsupervised Adaptation

TL;DR

This paper introduces a novel unsupervised learning framework for high-level vision tasks that enhances low-light images using a learnable, concave illumination curve inspired by camera response functions, improving performance across multiple vision applications.

Contribution

It proposes the first learnable, unsupervised illumination enhancement model based on a concave curve, enabling high-level vision tasks to adapt to low-light conditions without annotated data.

Findings

Outperforms existing low-light enhancement methods

Demonstrates superior generalization across various vision tasks

Effective in unsupervised domain adaptation for low-light scenarios

Abstract

Low light conditions not only degrade human visual experience, but also reduce the performance of downstream machine analytics. Although many works have been designed for low-light enhancement or domain adaptive machine analytics, the former considers less on high-level vision, while the latter neglects the potential of image-level signal adjustment. How to restore underexposed images/videos from the perspective of machine vision has long been overlooked. In this paper, we are the first to propose a learnable illumination enhancement model for high-level vision. Inspired by real camera response functions, we assume that the illumination enhancement function should be a concave curve, and propose to satisfy this concavity through discrete integral. With the intention of adapting illumination from the perspective of machine vision without task-specific annotated data, we design an asymmetric cross-domain self-supervised training strategy. Our model architecture and training designs mutually benefit each other, forming a powerful unsupervised normal-to-low light adaptation framework. Comprehensive experiments demonstrate that our method surpasses existing low-light enhancement and adaptation methods and shows superior generalization on various low-light vision tasks, including classification, detection, action recognition, and optical flow estimation. Project website: https://daooshee.github.io/SACC-Website/